Bulldozers first screens

[Dumo]

NDA made peep feels like a VIP. Lol

Just wait for retail chip...Its probly better overclock as well

[MAS]

zambezi result taken from here

[informal]

zambezi result taken from here

It's old rumor from rumorpedia.Dated back from February.

[cold2010]

AMD 900-Series Bulldozer Chipsets to Launch at Computex 2011

At this year's Computex expo, that is scheduled to take place between May 31 and June 4 in Taipei, Taiwan, motherboard makers will introduce their first AM3+ Bulldozer solutions based on the 900-series chipset logic developed by AMD.

According to sources form Taiwanese motherboard makers, AMD hasn't established an official launch date for Bulldozer, but allowed them to display their solution based on the 900-series chipset at the Computex fair.

“There are no exact embargo dates at the moment but the latest information from AMD is that we are allowed to launch at Computex,” said a source close to the industry that was quoted by the Sweclockers publication.

The 900-series will include three new motherboard chipsets, dubbed 990FX, 990X, and 970 and all of these support AM3+ processors as well as Nvidia's SLI technology.

The main difference between the three chipsets is the amount of PCI Express lanes available as the 990FX features 40 such lanes that can be routed to run two PCI Express x16 slots or four PCIe x8 slots, while the 990X packs just a pair of PCIe x8 slots.

The 990FX is compatible with two-way (2 x16) and three-way (3 x8) SLI setups.

All of the 900-series motherboard chipsets feature SATA 6Gbps with TRIM support, RAID 0/1/10/5, IOMMU I/O virtualization, as well as USB 3.0 support (via third party controllers).

The 900-series motherboards are also compatible with older AM3 processors.

All the reports that have been arriving until now stated that Bulldozer is scheduled for an early June release, so AMD could be considering a Computex launch for the Zambezi FX desktop chips.

AMD will initially launch four Zambezi FX CPUs, two featuring an eight-core design, while the other two will packs six and four processing cores respectively and additional chips are expected in the October-November timeframe

[undone]

zambezi result taken from here

According to this rumor, the 3.5Ghz bulldozer is a Quad-Core Zambezi, then it is way too optimistic.

[Mechanical Man]

According to this rumor, the 3.5Ghz bulldozer is a Quad-Core Zambezi, then it is way too optimistic.

Based on what?

[XRL8]

Based on what?

Reading the article i dont see it mentioning quadcore anywhere, besides lotsa people are going to made this mistake of module/core for some time.

[informal]

Article actually mentions it's supposedly a quad core 3.5Ghz Zambezi(read the first sentence of the article).Sorry but there is no way a quad core Zambezi can score 27K in vantage CPU test.On the other hand,an 8 core Zambezi with Turbo @ 4.2Ghz can score around 32K. That's close to the score of Westmere 6C or SB-E 6C.

[undone]

Based on what?

We have just received a photo of what sender claims to be an engineering sample of AMD Bulldozer (Zambezi) Quad Core processor benchmark score table.

I guess they mixed up the quad-module & quad-core.

[Mechanical Man]

Article actually mentions it's supposedly a quad core 3.5Ghz Zambezi(read the first sentence of the article).Sorry but there is no way a quad core Zambezi can score 27K in vantage CPU test.On the other hand,an 8 core Zambezi with Turbo @ 4.2Ghz can score around 32K. That's close to the score of Westmere 6C or SB-E 6C.

Based on what? I dont say it can, but i wonder how anyone knows it cant.

[undone]

Have this already been posted? If yes then please ignore me.

[informal]

Based on what? I dont say it can, but i wonder how anyone knows it cant.

Because it would need to have around 2.23x fp throughput per core and per clock Vs Deneb to do so.Which won't happen. In numbers ,Deneb QC @ 3Ghz scores 10.4K ,so it would take 2.23x the throughput of Deneb to score 23K with QC Zambezi @ 3Ghz and another ~16% clock speed to get to 27K.

Although Zambezi's fmacs are really powerful ,they are not more than 2x faster than one family 10h core in SSE(comparing one 128bit fmac with one Deneb core). Looking at AMD's own projections for spec fp rate and then looking at leaked donanimhaber slide with C11.5 score,we see a curious matching of performance increase in fp workload,of around 1.88x .This is the combined effect of 33% more cores,more fp execution resources and more clockspeed.Translated into IPC terms,around 1.3x or 30% improvement fmac vs core at similar clock. Note that this may vary a lot since we have no idea in which specific cases fmacs can excel and in which they may have lower throughput.

edit:
@undone
Nope it hasn't been posted and thanks! Looks nice . All the ISA extensions crammed into Orochi design need a wider CPUz window in order to be displayed properly .
Can you tell us from where did you get this image?

[XRL8]

Article actually mentions it's supposedly a quad core 3.5Ghz Zambezi(read the first sentence of the article).Sorry but there is no way a quad core Zambezi can score 27K in vantage CPU test.On the other hand,an 8 core Zambezi with Turbo @ 4.2Ghz can score around 32K. That's close to the score of Westmere 6C or SB-E 6C.

Well yes youre right, im tired and missed it somewhow.
Its still possible them or sender mixed up quad core/module thing.
As to the shot above, no it hasnt been posted here yet, but it doesnt bring anything to the table.

[Olivon]

Can you tell us from where did you get this image?

http://translate.google.com/translat...%2Findex6.html

On the Zambezi thread by 2good4you

[informal]

http://translate.google.com/translat...%2Findex6.html

On the Zambezi thread by 2good4you

Thanks .Maybe first post in this thread can be updated too.

[tifosi]

Have this already been posted? If yes then please ignore me.

http://img228.imageshack.us/img228/7...ldozerleak.jpg

I'm a very happy man as of this moment...
I know it could be interpreted in so many WRONG ways, but i couldn't care less

[Lanek]

Im surprised greenzilla have not make an article about it yet... this processor should be really good

[Morais]

I'm a very happy man as of this moment...
I know it could be interpreted in so many WRONG ways, but i couldn't care less

It would be awesome if its at stock settings and cpu-z got all values right

[BeepBeep2]

I like the backround in that leak...

Fits my taste

[FlanK3r]

Think, its fake?...L1 Data is right 4-way, not?

[LightSpeed]

is there a larger image of that pic BeepBeep? 1920 maybe? Nice background..

[[XC] Oj101]

The specification bar looks MUCH better now If that's true I'm absolutely loving the stock voltage

[BeepBeep2]

is there a larger image of that pic BeepBeep? 1920 maybe? Nice background..

With or without the logos, and or the new ATI logo? That one is really old.

[Opteron146]

Think, its fake?...L1 Data is right 4-way, not?

Yes, fake. Plus there have been already better leaked screenshots with the correct data, e.g. here:
http://we.pcinlife.com/thread-1657459-1-1.html

[LightSpeed]

With or without the logos, and or the new ATI logo? That one is really old.

Any would do

Yeah I figured its old since theres an ati logo. must be from the 58xx gen? nm that tho, any of them would do tbh, the more recent it is the better